Medical Marijuana Laws and Workplace Fatalities in the United States
D. Mark Anderson, Ph.D.*
Department of Agricultural Economics and Economics
Montana State University
Email: [email protected]
Daniel I. Rees, Ph.D.
Department of Economics
University of Colorado Denver
Email: [email protected]
Erdal Tekin, Ph.D.
School of Public Affairs
American University
Email: [email protected]
August 2017
Word count: 2,944
* Corresponding author. Address: Department of Agricultural Economics and Economics, Montana State University,
P.O. Box 172920, Bozeman, MT 59717-2920. Phone: 406-366-0921.
Medical Marijuana Laws and Workplace Fatalities in the United States
Importance: To date, 29 states and the District of Columbia have legalized the use of marijuana
for medicinal purposes. Although there is increasing concern that legalizing medical marijuana
will make workplaces more dangerous, little is known about the relationship between medical
marijuana laws (MMLs) and workplace fatalities.
Objective: To determine the association between legalizing medical marijuana and workplace
fatalities.
Design: Repeated cross-sectional data on workplace fatalities at the state-year level were
analyzed using a multivariate Poisson regression.
Setting: Workplace fatalities reported to the Bureau of Labor Statistics for the period 1992-2015.
Participants: All 50 states and the District of Columbia.
Exposures: The legalization of medical marijuana.
Main Outcomes and Measures: The number of workplace fatalities in a state and year.
Multivariate Poisson regression models were adjusted for state demographics, the unemployment
rate, the decriminalization of marijuana, the legalization of recreational marijuana, state fixed
effects, and year fixed effects.
Results: During the period under study, 19 states and the District of Columbia legalized medical
marijuana. Legalizing medical marijuana was associated with a 19.5% reduction in the expected
number of workplace fatalities among workers aged 25-44 (incident rate ratio [IRR], .805; 95%
CI, .662-.979). The association between legalizing medical marijuana and workplace fatalities
among workers aged 16-24, although negative, was not statistically significant at conventional
levels. Likewise, the association between legalization and workplace fatalities among workers
aged 45 and older was not statistically significant. The association between legalizing medical
marijuana and workplace fatalities among workers aged 25-44 grew stronger over time. Five
years after coming into effect, legalization was associated with a 41.1% reduction in the expected
number of workplace fatalities (IRR, .663; 95% CI, .482-.912). Medical marijuana laws that
listed pain as a qualifying condition or allowed collective cultivation were associated with larger
reductions in fatalities among workers aged 25-44 than those that did not.
Conclusions and Relevance: The results provide evidence that legalizing medical marijuana
improved workplace safety for workers aged 25-44. Further investigation is required to
determine whether this result is attributable to reductions in the consumption of alcohol and other
substances that impair cognitive function, memory, and motor skills.
1
INTRODUCTION
Although marijuana remains illegal under federal law, 29 states and the District of
Columbia have passed laws legalizing its use for medicinal purposes.1 Medical marijuana laws
(hereafter MMLs) remove state-level penalties for using and possessing marijuana for medical
purposes. Patients are required to obtain approval or certification from a physician, and
physicians who recommend marijuana to their patients are immune from criminal prosecution.
Increasingly, concerns are being raised over the potential impact of MMLs on workplace
safety.2-5 As a backdrop to these concerns, there are important legal issues surrounding
workplace safety and the use of medical marijuana that remain unresolved. For instance, in
many states it is unclear whether employers can impose sanctions on registered medical
marijuana patients who test positive for tetrahydrocannabinol (THC), the active ingredient in
marijuana, or whether insurance companies can claim marijuana as the cause of injury or death
in the workplace.6-8
There is strong evidence that legalizing medical marijuana leads to a decrease in the price
of marijuana and an increase in its consumption.9-11 However, the association between
legalization and workplace safety could, in theory, be negative or positive. On the one hand,
extensive research has demonstrated that there are important short-term effects of marijuana use
on psychomotor performance and cognition that could lead to more on-the-job accidents,
including impairments in memory function, information processing, hand-eye coordination, and
reaction times.12-15 On the other hand, previous studies have found that the legalization of
medical marijuana leads to substantial reductions in the consumption of alcohol, opioids and
other substances,9, 16-19, which could lead to safer workplaces and fewer accidents.
Drawing on data at the state-year level collected by the Bureau of Labor Statistics for the
2
period 1992-2015, the relationship between legalizing medical marijuana and workplace
fatalities was examined. Multivariate Poisson regression analysis was used to adjust for
demographics, income, the unemployment rate, legalization of recreational marijuana,
decriminalization of marijuana, state fixed effects, and year fixed effects. No previously
published study has examined the relationship between MMLs and workplace fatalities.
METHODS
Panel data on workplace fatalities at the state-year level came from Census of Fatal
Occupational Injuries (CFOI). These data are produced by the Occupational Safety and Health
Statistics (OSHS) program, which is administered by the Bureau of Labor Statistics (BLS). The
CFOI provides counts of all fatal work injuries occurring in the U.S. during each calendar year.
The OSHS program uses diverse state, federal, and independent data sources to identify, verify,
and describe fatal work injuries, ensuring that counts are as complete and accurate as possible.
The CFOI data are publicly available from the BLS for the period under study, 1992-2015, and
have been used by previous researchers interested in the determinants of workplace saftey.20-22
Total workplace fatality counts by state and year, as well as counts for different age groups are
available. Twenty-four years multiplied by 51 (50 states and the District of Columbia) yielded a
total of 1224 observations for analysis.
Information on the state-level legalization of medical marijuana is reported in Table 1.
During the period under study, 24 states and the District of Columbia adopted MMLs, although
medical marijuana programs were not yet operational in 4 of these states (Maryland, Minnesota,
New Hampshire, and New York). Sixteen of the remaining 20 states permitted patients to
register on the basis of pain, which could encourage recreational use as opposed to medical use
3
for severe or terminal illnesses.23 Twelve of the 20 states that legalized medical marijuana
during the period under study prohibited collective cultivation, also known as “group growing”,
either by limiting caregivers to 1 patient or restricting where and how marijuana can be grown.
Law enforcement authorities assert that there is more diversion from the medicinal market to the
recreational market in states that allow collective cultivation.24
Multivariate Poisson regression analysis was used to estimate the association between an
indicator (i.e., a 0/1 variable) of medical marijuana legalization and the expected number of
workplace fatalities per 100,000 workers in a given state and year. If a MML came into effect
after January 1, it was coded as a fraction for that year (e.g., it was coded as 0.5 if the law came
into effect on July 1). Incident rate ratios (IRRs) were considered statistically significant if their
95% confidence interval (CI) did not include the value of 1. We corrected standard errors
(which were used to calculate CIs and p-values) for clustering at the state level.25
Following previous studies in this area of research9-11, 16-17, 19, 50 state indicators were
included as covariates in the regression analysis. Their inclusion on the right-hand side of the
regression model accounted for the influence of time-invariant factors at the state level (i.e., state
“fixed effects”) such as rules and regulations pertaining to workplace safety and ensured that
estimates of the association between legalizing medical marijuana and workplace fatalities were
identified using only within-state variation over time. Again, following previous studies in this
area of research9-11, 16-17, 19, 23 year indicators were included to account for year-to-year changes
in workplace fatalities that were common across all 50 states and the District of Columbia due to,
for instance, changes in federal regulations or technology.
Estimated IRRs were also adjusted for within-state changes over time in demographics
(percent of the population white, percent black, and percent over the age of 18), per-capita
4
income, the unemployment rate, the legalization of recreational marijuana, and the
decriminalization of marijuana. There is evidence that workplace accidents are generally pro-
cyclical26, while the effects of legalizing recreational marijuana and decriminalizing marijuana
could be similar to the effects of legalizing medical marijuana. Information on MMLs, whether
the use of recreational marijuana was legal, and the decriminalization of marijuana came from a
variety of published sources.1, 27-28 The state unemployment rate and per-capita income came
from the Bureau of Labor Statistics and the Bureau of Economic Analysis, respectively.29-30
Means of workplace fatalities and the covariates are reported in Table 2.
Several extensions of the basic multivariate Poisson regression model described above
were also estimated. Specifically, the expected number of workplace fatalities was replaced by
the expected number of workplace fatalities for the following age groups: 16-24, 25-44, 45-64,
and 65+. Previous studies provide evidence that legalizing medical marijuana has larger effects
on marijuana consumption among young adults than among adolescents or older adults.10, 31-33 In
addition, regressions were run allowing the association between MMLs and workplace fatalities
to vary according to whether pain was included as a qualifying condition and whether collective
cultivation was allowed. Critics of legalizing medical marijuana contend that including chronic
or severe pain as a qualifying condition encourages recreational use by registered patients,23
while there is evidence that the effect of MMLs on substance use is strongest when collective
cultivation is allowed.9, 19, 34 Finally, estimates of the association between MMLs and workplace
fatalities were allowed to vary according to the time elapsed since implementation to account for
delays in patient registration and the opening of dispensaries. These estimates were used to
produce an event-study figure, which allowed an examination of pre-treatment trends in
workplace fatalities. There is evidence that the effect of MMLs on substance use is weakest
5
immediately after implementation, becoming steadily stronger thereafter.9-10, 16
RESULTS
The association between legalizing medical marijuana and total workplace fatalities,
although negative, was not statistically significant at conventional levels (Table 3). By contrast,
legalizing medical marijuana was associated with a 19.5% reduction in the expected number of
workplace fatalities among workers aged 25-44 (IRR, .805; 95% CI, .662-.979) after adjusting
for the covariates listed in Table 2, state fixed effects, and year fixed effects. This negative
association was robust to including state-specific linear trends in the model. The association
between MMLs and workplace fatalities among workers aged 16-24, although negative, was not
statistically significant. Likewise, the association between MMLs and workplace fatalities
among workers aged 45 and over was negative but not statistically significant at the 5% level.
The negative association between legalizing medical marijuana and workplace fatalities
among workers aged 25-44 was strongest if pain was included as a qualifying condition (Table 4,
Panel I). Specifically, legalizing medical marijuana was associated with a 19.8% reduction in
the expected number of workplace fatalities among workers aged 25-44 (IRR, .802; 95% CI,
.657-.978) if pain was included as a qualifying condition; if pain was not included as a qualifying
condition, the association between legalizing medical marijuana and workplace fatalities was not
statistically significant (IRR, .980; 95% CI, .859-1.12).
The negative association between MMLs and workplace fatalities among workers aged
25-44 was stronger in states that allowed collective cultivation than in states that did not (Table
4, Panel II). Specifically, legalizing medical marijuana was associated with a 20.5% reduction in
the expected number of workplace fatalities among workers aged 25-44 (IRR, .795; 95% CI,
6
.651-.971) if collective cultivation was allowed; if collective cultivation was not allowed, the
association between legalizing medical marijuana and workplace fatalities was not statistically
significant (IRR, 1.03; 95% CI, .832-1.27).
Prior to the year of implementation (year 0), there was no evidence of an association
between MMLs and workplace fatalities among workers aged 25-44 after adjusting for the
covariates listed in Table 2, state fixed effects, and year fixed effects (Figure 1). After
implementation, the negative association between MMLs and workplace fatalities among
workers aged 25-44 gained strength over time. In the first 4 years after implementation, the
association between legalization and workplace fatalities was relatively small and not
statistically significant. For instance, one year after implementation the IRR was 1.01 [95% CI,
.903-1.14], and three years after implementation the IRR was .861 [95% CI, .741-1.00].
However, 5 or more years after implementation, the association between MMLs and workplace
fatalities was negative, larger in absolute magnitude, and statistically significant. Specifically,
legalizing medical marijuana was associated with a 33.7% reduction in the expected number of
workplace fatalities among workers aged 25-44 (IRR, .663; 95% CI, .482-.912).
DISCUSSION
MMLs protect patients from criminal prosecution, but the use of medical marijuana in the
workplace is generally not protected. Several states (e.g., Colorado, Michigan, Oregon,
Washington) have ruled that employers may discipline employees or terminate their employment
following a positive drug test because MMLs are not intended to address employment issues.6-7
Although some MMLs explicitly protect employees from termination due to a positive drug test,
employers and insurance companies continue to argue in court that the use of medical marijuana
7
violates zero-tolerance drug policies and compromises workplace safety.4, 6-8
Unresolved legal issues notwithstanding, it is clear that workplace injuries impose
substantial costs on society. In 2015, the latest year for which statistics are available, there were
a total of 4,836 fatal on-the-job injuries in the United States.35 The causes of these injuries are
myriad, ranging from homicide, to motor vehicle accidents, to electrocution, to being crushed by
machinery.35 Given their frequency, any estimate of the cost of legalizing medical marijuana is
likely to be inaccurate if workplace fatalities are not taken into account.
Legalizing the use of medical marijuana should unambiguously lead to an increase in the
consumption of marijuana.36-37 By contrast, the association between legalization and workplace
safety could, in theory, be either negative or positive. Marijuana use impairs memory function,
information processing, hand-eye coordination, and reaction times12-15, all of which could
plausibly result in more on-the-job accidents and workplace fatalities. Indeed, a number of
epidemiological studies provide evidence of a positive association between marijuana use and
the likelihood of being involved in a motor vehicle accident38, one of the leading causes of on-
the-job injury in the United States.35
However, other studies show that the legalization of medical marijuana is associated with
substantial reductions in the consumption of alcohol, opioids and other substances.9, 16-19 For
instance, Anderson et al. found that legalization of medical marijuana was associated with a 5%
reduction in beer sales9, while Bachhuber et al. found that legalization of medical marijuana was
associated with a 20% to 33% decrease in deaths involving opioids.16 Because the use of alcohol
at work is associated with a substantial increase in the risk of injury39-40, and because non-
habitual opioid use slows reflexes and impairs cognitive functioning41, the enactment of MMLs
could, in theory, make workplaces safer.
8
Traffic fatalities, the abuse of other substances, suicides, and crime are among the costs
of legalizing medical marijuana considered by previous researchers.9, 11, 16, 42-45 The current
study, however, is the first to examine the association between MMLs and workplace fatalities.
Using data from the BLS and a multivariate Poisson regression model, a negative association
between MMLs and fatalities among workers aged 25-44 was found after adjusting for state
demographics, the unemployment rate, the decriminalization of marijuana, the legalization of
recreational marijuana, state fixed effects, and year fixed effects.
The negative association between MMLs and workplace fatalities among workers aged
25-44 was robust to including state-specific linear trends in the model, suggesting that it cannot
be explained by slowly evolving, but difficult-to-measure factors at the state level such as
attitudes or health behaviors.16 The association between MMLs and workplace fatalities among
workers aged 16-24, although negative, was not statistically significant. Likewise, the
association between MMLs and workplace fatalities among worker above the age of 44 was not
statistically significant. This general pattern of results is not surprising given that previously
published research suggests that young adults respond to the legalization of medical marijuana
by consuming substantially less alcohol,9, 19 while any response to legalizing medical marijuana
on the part of teenagers and older adults has been difficult to isolate from year-to-year
fluctuations in substance use.9-10, 19, 31-33
Two studies provide evidence that MMLs passed in the 1990s and early 2000s led to
higher enrollment rates and greater marijuana consumption than did newer “medicalized”
medical marijuana programs.23, 46 Consistent with these findings, the negative association
between MMLs and workplace fatalities among adults aged 25-44 was stronger in states that
included severe or chronic pain as a qualifying condition. Consistent with the argument that
9
diversion to the recreational market can be limited by prohibiting home growing and limiting
caregivers to one patient24, the negative association between MMLs and workplace fatalities
among adults aged 25-44 was stronger in states that allowed collective cultivation. Finally,
consistent with the parallel trends assumption, there was no evidence of any association between
MMLs and workplace fatalities among workers aged 25-44 in the years leading up to
implementation. However, the negative association between MMLs and workplace fatalities
gradually became larger in magnitude after implementation.
This study has several limitations that deserve mention. First, the data are at the state as
opposed to the individual level. Access to detailed individual-level data collected by the BLS
would allow us to explore who precisely was affected by MMLs and would allow us to describe
the nature of their injuries. Access to individual-level data could also improve the precision of
the estimates and would allow us to explore the association between MMLs and workplace
fatalities in specific industries or occupations by age (e.g., young adults working in mining,
young adults working in construction, etc.). The negative associations between MMLs and
workplace fatalities found for other age groups (i.e., workers aged 16-24, workers aged 45-64,
and workers over the age of 64) were not statistically significant. Using individual-level data
would allow us to adjust for other factors (e.g., marital status, educational attainment,
occupation) and could produce more precise estimates of these associations.
Second, the estimates provided in Tables 3 and 4 demonstrate a negative association
between MMLs and workplace fatalities among adults aged 25-44, but it is not clear why this
negative association exists. This negative association could be the result of workers responding
to the legalization of medical marijuana by reducing their use of other substances (e.g., alcohol
and prescription opioids). It is also possible that this negative association is due to other, often
10
more-difficult-to-document, effects of THC. For instance, drivers under the influence of THC
appear to take fewer risks2, which could reduce fatal accidents among, for instance, truck drivers
and other transportation workers. Consistent with the argument that workers are healthier, and
perhaps less prone to being involved in an accident, the legalization of medical marijuana has
been linked to fewer sickness-related absences from work.47 In the absence of detailed data on
drug and alcohol use by workers involved in accidents, it will be difficult to distinguish between
these various potential mechanisms.
Third, and finally, the inclusion of state fixed effects in our regression models accounted
for the influence of time-invariant factors that could be correlated with MMLs and workplace
fatalities. However, it is possible that there were difficult-to-observe factors at the state level
that changed over time and led both to the adoption of MMLs and to reductions in workplace
fatalities. Although there was no evidence of reductions in workplace fatalities prior to the
adoption of MMLs (Figure 1), if these factors changed at the same time as the legalization of
medical marijuana, their influence would have been difficult to detect.
CONCLUSION
Although 29 states and the District of Columbia have passed laws legalizing marijuana
for medicinal use, there is a dearth of evidence on the relationship between medical marijuana
and workplace safety. The current study is the first to explore the effects of medical marijuana
laws on workplace fatalities. Our results suggest that legalizing medical marijuana leads to a
reduction in workplace fatalities among workers aged 25-44. This reduction may be the result of
workers substituting marijuana in place of alcohol and other substances that can impair cognitive
function and motor skills; however, it is important to note that we cannot rule out other potential
11
mechanisms. As the debate over legalization of medical and recreational marijuana continues, it
is important that we learn more about the effects of MMLs on workplace safety.
12
ACKNOWLEDGEMENTS
Author Contributions: Drs. Anderson and Rees had full access to the data in the study and take
responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Anderson, Rees, Tekin
Acquisition, analysis, or interpretation of data: Anderson, Rees
Drafting of the manuscript: Anderson, Rees, Tekin
Critical revision of the manuscript for important intellectual content: Anderson, Rees, Tekin
Statistical analysis: Anderson
Obtained funding: Anderson
Administrative, technical, or material support: Anderson
Study supervision: Anderson, Rees, Tekin
Conflict of Interest Disclosures: None reported.
Funding/Support: Partial support for this research came from a Eunice Kennedy Shriver
National Institute of Child Health and Human Development research infrastructure grant, R24
HD042828, to the Center for Studies in Demography and Ecology at the University of
Washington.
Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study;
collection, management, and interpretation of the data; preparation, review, or approval of the
manuscript; and decision to submit the manuscript for publication.
13
REFERENCES
1. State Medical Marijuana Laws. National Conference of State Legislatures.
http://www.ncsl.org/research/health/state-medical-marijuana-laws.aspx. Updated July 7, 2017.
Accessed July 28, 2017.
2. Goldsmith RS, Targino MC, Fanciullo GJ, Martin DW, Hartenbaum NP, White JM, Franklin
P. Medical marijuana in the workplace challenges and management options for occupational
physicians. J Occup Environ Med. 2015; 57(5):518–525.
3. Parnes JE, Bravo AJ, Conner BT, Pearson MR. A burning problem: cannabis lessons learned
from Colorado. Addict Res Theory. In Press.
4. Phillips JA, Holland MG, Baldwin DD, Gifford-Meuleveld L, Mueller KL, Perkison B, Upfal
M, Dreger M. Marijuana in the workplace: guidance for occupational health professionals and
employers: joint guidance statement of the American Association of Occupational Health Nurses
and the American College of Occupational and Environmental Medicine. Workplace Health Saf.
2015; 63(4):139-64.
5. Schwartz ND. Economy needs workers, but drug tests take a toll. New York Times.
https://www.nytimes.com/2017/07/24/business/economy/drug-test-labor-hiring.html?_r=0.
Published July 24, 2017. Accessed July 28, 2016.
6. Deitchler, DL. The workplace and medical marijuana: employer guidelines for navigating the
legal haze. Inside Counsel. http://www.insidecounsel.com/2015/04/15/the-workplace-and-
medical-marijuana-employer-guide. Updated April 15, 2015. Published July 28, 2017.
7. Hickox, SA. Drug testing of medical marijuana users in the workplace: an inaccurate test of
impairment. Hofstra Lab & Emp LJ. 2012; 29(2):273-341.
8. Olafson JJ. Marijuana use and the workplace. Employ Relat Today. 2016; 43(3):67–97.
9. Anderson DM, Hansen B, Rees DI. Medical marijuana laws, traffic fatalities, and alcohol
consumption. J Law Econ. 2013; 56(2):333-369.
10. Chu, Y. The effects of medical marijuana laws on illegal marijuana use. J Health Econ.
2014; 38(December):43-61.
11. Chu, Y. Do medical marijuana laws increase hard-drug use? J Law Econ. 2015; 58(2):481-
517.
12. Hall W. Adverse health effects of non-medical cannabis use. Lancet. 2009; 374(9698):1383-
1391.
13. Hartman RL, Huestis MA. Cannabis effects on driving skills. Clin Chem. 2013; 59(3):478-
492.
14
14. Marijuana. National Institute on Drug Abuse.
https://www.drugabuse.gov/publications/drugfacts/marijuana. Published February 28, 2017.
Accessed July 28, 2017.
15. Ramaekers JG, Berghaus G, van Laar M, Drummer OH. Dose related risk of motor vehicle
crashes after cannabis use. Drug Alcohol Depend. 2004; 73(2):109–119.
16. Bachhuber MA, Saloner B, Cunningham CO, Barry CL. Medical cannabis laws and opioid
analgesic overdose mortality in the United States, 1999-2010. JAMA Intern Med. 2014; 174(10):
1668-1673.
17. Bradford AC, Bradford WD. Medical marijuana laws reduce prescription medication use in
Medicare Part D. Health Affairs. 2016; 35(7):1230-1236.
18. Reiman A. Cannabis as a substitute for alcohol and other drugs. Harm Reduct J. 2009; 6:35.
19. Sabia JJ, Swigert J, Young T. The effect of medical marijuana laws on body weight. Health
Econ. 2017; 26(1):6-34.
20. Morantz AD. Has devolution injured American workers? State and federal enforcement of
construction safety. J Law Econ Organ. 2009; 25(1):183-210.
21. Smith SM. Workplace hazards of truck drivers. Mon Labor Rev, U.S. Bureau of Labor
Statistics. https://doi.org/10.21916/mlr.2015.8. Published April 2015. Accessed July 20, 2017.
22. Mendeloff J, Staetsky L. Occupational fatality risks in the United States and the United
Kingdom. Am J Ind Med. 2014; 57(1):4–14.
23. Williams AR, Olfson M, Kim JH, Martins SS, Kleber HD. Older, less regulated medical
marijuana programs have much greater enrollment rates than newer 'medicalized' programs.
Health Aff. 2016; 35(3):480-488.
24. Selecky M. Patient Access to Medical Marijuana in Washington State. Publication No. 631-
001. Olympia: Washington State Department of Health.
http://www.doh.wa.gov/portals/1/Documents/2000/PatientAccess.pdf. Published July 2008.
Accessed July 28, 2017.
25. Bertrand M, Duflo E, Mullainathan S. How much should we trust differences-in-differences
estimates? Q J Econ. 2004; 119(1):249-276.
26. Boone J, van Ours JC, Wuellrich J, Zweimüller J. Recessions are bad for workplace safety. J
Health Econ. 2011; 30(4):764-773.
27. Scott E. Marijuana Decriminalization. https://www.cga.ct.gov/2010/rpt/2010-R-0204.htm.
Updated May 5, 2010. Accessed July 24, 2017.
15
28. Marijuana Overview. National Conference of State Legislatures.
http://www.ncsl.org/research/civil-and-criminal-justice/marijuana-overview.aspx. Updated April
3, 2017. Accessed July 24, 2017.
29. Local area unemployment statistics. Bureau of Labor Statistics, United States Department of
Labor. http://www.bls.gov/lau. Accessed July 15, 2017.
30. Regional economic accounts. Bureau of Economic Analysis, United States Department of
Commerce. http://www.bea.gov/regional/index.htm. Accessed July 15, 2017.
31. Anderson DM, Hansen D, Rees DI. Medical marijuana laws and teen marijuana use. Am Law
Econ Rev. 2015; 17(2):495-528.
32. Hasin DS, Wall M., Keyes MK, Cerdá M, Schulenberg J, O'Malley PM, Sandro G, Pacula
R, Feng T. Medical marijuana laws and adolescent marijuana use in the USA from 1991 to 2014:
results from annual, repeated cross-sectional surveys. Lancet Psychiat. 2015; 2(7):601-608.
33. Lynne-Landsman SD, Livingston MD, Wagenaar AC. Effects of state medical marijuana
laws on adolescent marijuana use. Am J Public Health. 2013; 103(8):1500–1506.
34. Smart R. The kids aren’t alright but older adults are just fine: effects of medical marijuana
market growth on substance use and abuse. Available at SSRN:
http://dx.doi.org/10.2139/ssrn.2574915. Published November 25, 2015. Accessed July 31, 2017.
35. Census of Fatal Occupational Injuries Summary, 2015. Bureau of Labor Statistics, United
States Department of Labor. https://www.bls.gov/news.release/cfoi.nr0.htm. Published
December 16, 2016. Accessed August 1, 2017.
36. Anderson DM, Rees DI. The legalization of recreational marijuana: how likely is the worst-
case scenario? J Policy Anal Manag. 2014; 33(1):221-232.
37. Pacula R., Chriqui J, Reichmann D, Yvonne T. State medical marijuana laws: understanding
the laws and their limitations. J Public Health Policy. 2002; 23(4):413-439.
38. Asbridge M, Hayden JA, Cartwright JL. Acute cannabis consumption and motor vehicle
collision risk: systematic review of observational studies and meta-analysis. BMJ. 2012; 344:
e536.
39. Ohsfeldt RL, Morrisey MA. Beer taxes, workers’ compensation, and industrial injury. Rev
Econ Statistics. 1997; 79(1):155-160.
40. Ramirez M, Bedford R, Sullivan R., Anthony TR, Kraemer J, Faine B., Peek-Asa C.
Toxicology testing in fatally injured workers: a review of five years of Iowa FACE cases. Int J
Environ Res Public Health, 2013; 10(11):6154-6168.
16
41. Zacny, JP. A review of the effects of opioids on psychomotor and cognitive functioning in
humans. Exp Clin Psychopharm. 1995; 3(4):432-466.
42. Anderson, DM, Rees DI, Sabia JJ. Medical marijuana laws and suicides by gender and age.
Am J Public Health. 2014; 104(12):2369-2376.
43. Morris RG, Teneyck M, Barnes JC, Kovandzic, TV. The effect of medical marijuana laws on
crime: evidence from state panel data, 1990-2006. PloS ONE, 2014; 9(3):e92816.
44. Salomonsen-Sautel S., Min S., Sakai, JT, Thurstone C, Hopfera, C. Trends in fatal motor
vehicle crashes before and after marijuana commercialization in Colorado. Drug Alcohol
Depend. 2014; 140(1 July):137–144.
45. Wen H., Hockenberry J, Cummings JR., The effect of medical marijuana laws on adolescent
and adult use of marijuana, alcohol, and other substances. J Health Econ. 2015; 42(July):64-80.
46. Williams AR, Santaella‐Tenorio J, Mauro CM, Levin FR, Martin SS. Loose regulation of
medical marijuana programs associated with higher rates of adult marijuana use but not cannabis
use disorder. Addiction. In Press.
47. Ullman DF. The effect of medical marijuana on sickness absence. Health Econ. In Press.
17
.4.6
.81
1.2
-3 -2 -1 0 1 2 3 4 +5
Years Since Medical Marijuana Law Came Into Effect
Notes: Incident rate ratios (and their 95% confidence intervals) from a Poisson regression are reported. Thedependent variable is equal to the expected number of workplace fatalities among 25-44 year-olds in state s andyear t. Estimates are adjusted for the covariates listed in Table 2, 50 state indicators, and 23 year indicators.
Figure 1. Trends in Workplace Fatalities among 25-44 Year-Olds
18
Table 1. Medical Marijuana Laws 1992-2015
Effective date
Pain listed as
qualifying
condition
Collective
cultivation
allowed
Alaska March 4, 1999 Yes No
Arizona April 14, 2011 Yes Yes
California November 6, 1996 Yes Yes
Colorado June 1, 2001 Yes Yes
Connecticut August 20, 2014a No No
Delaware June 26, 2015a Yes No
D. C. July 30, 2013a No No
Hawaii December 28, 2000 Yes No
Illinois November 9, 2015a No No
Maine December 22, 1999 Yes No
Maryland Passed but not operationalb
Massachusetts January 1, 2013 No No
Michigan December 4, 2008 Yes Yes
Minnesota Passed but not operationalb
Montana November 2, 2004 Yes Noc
Nevada October 1, 2001 Yes Yes
New Hampshire Passed but not operationalb
New Jersey December 6, 2012a Yes No
New Mexico July 1, 2007 No No
New York Passed but not operationalb
Oregon December 3, 1998 Yes Yes
Rhode Island January 3, 2006 Yes Yes
Vermont July 1, 2004 Yesd No
Washington November 3, 1998 Yes Yes a Date on which first medical marijuana dispensary opened.
b MML passed during period 1992-2015, but first dispensary did not open until after 2015.
c Prior to Senate Bill 423 (July 1, 2011), Montana allowed for collective cultivation.
d “Pain” added to list of qualifying conditions in 2007.
19
Table 2. Descriptive Statisticsa
Mean
(SD)
Definition
Outcomes
Total Fatalities 108.3
(105.3)
Total workplace fatalities
Age 16-24 Fatalities 8.99
(11.9)
Workplace fatalities among 16-24-year-olds
Age 25-44 Fatalities 45.0
(49.2)
Workplace fatalities among 25-44-year-olds
Age 45-64 Fatalities 41.5
(39.2)
Workplace fatalities among 45-64-year-olds
Age 65+ Fatalities 10.6
(9.88)
Workplace fatalities among 65+ year-olds
State-level covariates
MML .158
(.362)
= 1 if state legalized marijuana for medicinal
use, = 0 otherwise
RML .003
(.052)
= 1 if state has legalized marijuana for
recreational use, = 0 otherwise
Decriminalization .243
(.428)
= 1 if state decriminalized marijuana, = 0
otherwise
Unemployment 5.69
(1.89)
State unemployment rate
Income 37,817
(7,188)
State real income per capita (2010 dollars)
% White .822
(.139)
Percent of state population white
% Black .117
(.116)
Percent of state population black
% Adult .752
(.023)
Percent of state population 18+ years of age
N = 1,224 a Unweighted means (with standard deviations in parentheses) are reported.
20
Table 3. Estimates of Relationship between Legalizing Medical Marijuana and
Workplace Fatalitiesa
Total
Fatalities
Ages 16-24
Fatalities
Ages 25-44
Fatalities
Ages 45-64
Fatalities
Ages 65+
Fatalities
MML .868
[.712, 1.06]
.946
[.786, 1.14] .805
[.662, .979]
.894
[.728, 1.10]
.986
[.746, 1.30] Boldface indicates statistical significance (p < 0.05)
a Each cell reports an incidence rate ratio (IRR) from a separate Poisson regression based on data
from the Bureau of Labor Statistics (1992-2015). All regressions are weighted by state employment
for the relevant population. IRRs are adjusted for the state characteristics listed in Table 2, 50 state
indicators, and 23 year indicators. Ninety-five % confidence intervals are in brackets. N = 1,224.
21
Table 4. Medical Marijuana Laws and Workplace Fatalities among 25-44-
Year-Olds by Whether Pain was a Qualifying Condition and Whether
Collective Cultivation was Alloweda
Panel I
Ages 25-44 Fatalities
MML – Pain Qualifying Condition
MML – Pain Not Qualifying Condition
.802
[.657, .978]
.980
[.859, 1.12]
Panel II
Ages 25-44 Fatalities
MML – Collective Cultivation Allowed
MML – Collective Cultivation Prohibited
.795
[.651, .971]
1.03
[.832, 1.27] Boldface indicates statistical significance (p < 0.05)
a Each panel reports IRRs from a separate Poisson regression based on data from the Bureau of
Labor Statistics (1992-2015). All regressions are weighted by state employment for the
population of 25-44-year-olds. IRRs are adjusted for the state characteristics listed in Table 2,
50 state indicators, and 23 year indicators. Ninety-five % confidence intervals are in brackets.
N = 1,224.